Welcome to What the Duck?! A podcast with real experts talking about direct spend challenges and experiences. And now, here’s your host, SourceDay’s very own manufacturing Maven, Sarah Scudder. Thanks for joining me for What the Duck?! Another Supply Chain Podcast brought to you by SourceDay. I’m your host, Sarah Scudder, and this is the podcast for people working in the direct materials part of the supply chain. I’m @SarahScudder on LinkedIn and @Sscudder on Twitter. Today, I’m going to be joined by Greg Manhire, and we’re going to discuss trying to manage Direct procurement for a global reverse logistics materials supply chain and fulfillment without an ERP system. Quite a mouthful, but we’ll dive into specifically what that means and how Greg was able to solve for that. So, if you work for a manufacturer and are struggling to properly track inventory consumption and replenishment, then this episode is for you. Greg enjoys working with small teams and suppliers to develop solutions aimed at improving the overall operation. He moved to Austin two years ago, just about the same time as me. I moved at the end of 2021, moved from Cali with his wife and teenage kids. He enjoys golfing and exercising and says that he needs to do more of both. Welcome to the show, Greg. Thanks, Sarah. It’s great to be here. So, why golf?

I mean, after being there a few years, it was growing at 100% a year. So, it was like, “Here’s a raise, and here’s a bonus, and here’s more stock.” And the people were fantastic. And the next thing you know, it made the top 10 places to work. Then it was top five, and then it was the number one place to work for a few years. So, when you’re in Silicon Valley and you’re kind of looking around going, “Where else can I go that’s better than this?” When you’re number one, there was nowhere else. And so, when you’re at a company that continued to be in the top 10 best places to work, it made it very easy to stay. But then, as it kind of fell out of that and Cloud started taking off, it definitely made it—I’d say it became more difficult to sell because customers were confused on really what they wanted. And so, it became a whole cost-cutting mode instead of a growth mode, which is a whole different experience, and culture changed, and it was that it was time to go.

Let’s talk a little bit about your time at NetApp, because you did some really interesting things that relate to direct materials procurement, which is the audience for this podcast. So, you started as something called a supply-based manager, and in this role, you were responsible for coordinating and implementing commodity strategies. What would you say was the most difficult commodity to manage at that time and why? I know you probably have to think long and hard because that was several years ago, yeah, more than several, but yeah, I mean, what I came in there and started with was storage, and the storage commodity, I’d say, evolved over time as we grew. So, as we grew, it became more specific, and I ended up focused on just disk drives. And so, I was the only person managing disk drives. We were growing from, you know, 30,000 a quarter or 50,000 to 100,000 disk drives a quarter. And it just, it became the highest spend commodity, so it was a highly visible commodity. So, there was no hiding when something went wrong. It was very visible. We were negotiating prices every quarter. And so, when you are the highest spend, you have the most focus on, “How are we doing on price reductions?” And on top of that, then we had to do QBRs every single quarter to stay on top of the relationship, the executive relationship, making sure the engineers are getting what they need, the roadmaps are aligning with where we’re headed. And then, as we started growing it, one supplier wasn’t enough. Then we needed two. So, we had to qualify a second supplier in the disk drive world. That’s not just throwing another disk drive in there. We had highly specific firmware requirements that they needed to meet, the qualified quality performance had to be the same.

So, when you look at all that, it was, it became, it was very, it’s a very, very technical commodity. But at the same time, it was just, it was challenging because of the volume and the scope of suppliers. And eventually, we brought on another commodity manager to help me, and then it was like, “Okay, got a little bit of breath.” But then we just kept growing, and the parts and the types of drives we were buying kept expanding. So, it just got more and more scrutiny as the spend went up and up and up.

What would you say of all the commodities that you worked on there, in what strategy turned out to be most impactful? Yeah, it was the disk drive strategy. Our manager at the time was pushing the commodity strategy concept, and so we became the guinea pigs because we had the highest spend and the most volume, and we became the guinea pigs that put together the strategy. So, I think it was most impactful for NetApp because it got that process off the ground and kind of became the gold standard of how other commodities should model their strategy after. But I think beyond that, it was it really helped our commodity team come together. So, we spent countless meetings and months prepping for the first strategy, aligning on, “Okay, here are all the topics we need to cover. Do we have it all covered?” And as we did it, you know, the team really jelled because you have a lot of different personalities with engineers and marketing folks and supply chain folks. We had a great team with a lot of great people, so it was enjoyable putting it together, spending time with them. But at the end of the day, we all came together and put together a strategy. And then, we walked away with everybody on the same page. We knew, “Okay, which suppliers are we working with? What products are we gonna go qualify? What are the volumes? Where do we think the pricing should be? And what are the issues for all the suppliers?” And that would translate into our QBRs as well. So, everybody would go into a QBR completely on the same page. We knew what the engineering issues were, we knew what the quality issues were, they knew what our supply chain issues were, and everybody backed each other up.

I mean, this doesn’t happen overnight. It’s a process, and it was highly coordinated, so it took a lot of time and effort. But at the end of the day, when I look back at those QBRs and those strategies, I think it was probably the most in sync I ever was with other organizations and teams, and aligned with where we were headed and why we were headed in those directions.

So, you were promoted to a Senior Supply-Based Manager, and in that role, you were responsible for reducing lead times for customer shipments. Why? Let’s start off with why this was so difficult.

Yeah, well, this was a new role that we had created because we went big on the on-track manufacturing model. We shut down our factory. We were probably one of the last factories actually building anything in Silicon Valley and went full bore with the contract manufacturing model. So, in this role, I essentially moved out of commodity management and went into managing our factory and managing the operation. We had very tight lead times already. I think we were at four days or three days for data storage systems, which are, if you’re not familiar, refrigerator-sized or even bigger cabinets filled with storage shelves or sheet metal boxes full of drives. Some people call them J-BODs (just a bunch of discs) and servers and switches and other components, and everything was configured to order. So, we had to have the process completely aligned so you could start at one end and say, “Okay, this one needs two storage shelves, this one needs 50 storage shelves,” and you could get that done and then move it to the next station. It was very much an assembly. And we put these things together, run the test. It was challenging trying to get that lead time down to three days, and it was all about metrics and how were we handling all these different areas. The commodity managers would get very frustrated with me when there were shortages because now I could see how their shortages impacted production, and since I knew their world, I could go back in and reach in and say, “Have you done this? Have you done this? Have you done this?” So, it became very annoyed with me being in their shorts, but it was challenging to keep the lead times where they were. And we had about a 97% on-time delivery to our three-day lead time, and if we had extended lead times, we would extend the lead times. But it was all about trying to stay on top of that and be proactive. So, when a customer did place an order, they knew they were going to get it in five days or ten days, and being able to deliver those dates was key because of the prep they’d have to do to be set up to receive 20 refrigerator-sized boxes at a time. A lot of coordination.

What would you attribute your success at being able to keep such narrow lead times? If you had to pick a couple of things that were really essential, I think the biggest thing that I see is flexibility and supply flexibility. We definitely would have shortages.

I mean, that happened. We would have parts fail on the line, but there was always another one in the box. If something would fail, you just get another one and put it in. We ran with just-in-time type inventory models, kanban models, so we always had product coming in the door. We were trying to stage. We looked at make first, buy, so we decided where was the most value added and where were we adding our value. So, we would try to push the value back to suppliers or I said the non-value-add items that we weren’t adding much value, push those back up the supply chain as far as we could. So, we’re really taking in assemblies into our contract manufacturers, putting those assemblies right into the systems, moving them into test within a day or two. I think that was key. And then running with a buffered supply model. Our management just said, “We are not going to miss a quarter-end.” And when you were ramping and growing at, you know, 100% quarter over quarter, then eventually that slowed down after we hit a billion, but you’re still running at 20%, 30%, 40% growth rates, it’s unpredictable. So, we always put on a 25% supply adder to all of our purchases. Was it expensive? It was expensive, but did we miss it? We never missed quarter-end because we were short on parts. So, I mean, and we were able to maintain those lead times because we had parts there. Those are some of the key things.

And I will point out that not all manufacturers are in a situation, a cash flow situation, to be able to order an extra 25%. If you’re working at a larger company, different situation than a smaller mid-size manufacturer. Yeah, and you know, it’s always going to depend on your growth too. I mean, I think one of the things that drove that number was the variability of our model. With the configure-to-order process, you’re not quite sure what people are going to order, right? And we had the forecasting pretty dialed in, but we had many different variants, very different variants of disk drives, NIC cards, HBA cards, extra cards you could add on if you wanted, fiber cables, copper cables. I mean, the options are sort of limitless, but there are many. So, you did have to have a lot of extra inventory around in case all of a sudden every order came in wanting fiber cables or every order wanted to be maxed out on HBAs, for example.

Greg, when you were explaining your process, you mentioned something that’s new to me. It’s called kanban. I would love to have you—I’m assuming it’s a fulfillment method— if you could kind of explain what that is and why and when or why would a manufacturer use it. Yeah, yeah. So, kanban is—I did a little research this morning just to make sure I wasn’t speaking out of turn. I know you have a lot of listeners and didn’t want any hate mail that that’s not right.

So, kanban, going back to the Japanese word, translates to a visual card, and it’s an aspect of lean manufacturing. So, when you look at it from an inventory management perspective, it basically translates into visual cues to signal replenishment. So, when we started off with this model, we did have a manufacturer who was about two hours away, and we were basically doing melt runs of parts every day or every two days. I can’t remember. But we had laminated cards, like three by five cards, with part numbers on them and a scannable label. And if we had four pallets on-site of that product, we had four cards on the shelf. One, if one pallet got moved to the manufacturing floor, they’d take the Velcro card off the shelf and they’d drop it in the bucket. And the on-site supplier rep would come downstairs every morning, see what got dumped in from the day before, take the cards, make a call to the factory and say, “Okay, ship this, this, this, and this.” And that afternoon or the next morning, it showed up, and we were replenishing our inventory.

So, I think, when does this make sense to use? I think you need to have some sort of standard product, right? Having standard parts that are on a consistent run rate definitely helps. We knew that based on our forecasts, like we would use like eight pallets a week, so we would put four in because that’s what we had, and we knew what our replenishment time was, that we could get one, two, three, or four replenished within 24 hours. We also had similar stocking locations at our supplier, so if we pulled all four, they could put four on a truck and have them down the next day. So, I think you need to understand what your capacity is or your run rate to be able to implement something like that. It definitely helps if you—you also need to understand your fulfillment lead time. If it takes you seven days to get something from your supplier, you need to know, “How much am I going to use in seven days? What’s the max I could use in seven days?” And then you need to plan out what those numbers are.

Our process evolved from Velcro cards on the pallet racks to more numbers in a system and to replenishment triggers. If we fell below this number, we would make more and we would track that. When I was managing the senior Global Supply manager leading that team, we had it all set up.

So, we had targets, and we were measuring every week at what level they were at throughout the quarter. So, it evolves over time, but I think the standardization of what you’re using, and it can, if you’re using it every day, every week, I think it makes sense. When you get into some low runners, it may be better at that point just to build as you need it because otherwise, now you’re just holding inventory and it’s just sitting there. And if you’re utilizing, in our case, it was assemblies, so if you’re not really using it, if it’s just sitting there, now you’re using components that could be used in your high-volume product. And we did come up with different strategies of build-to-order versus kanban-type product. I’ll have to ask some of our other guests if this is a strategy that they use because again, it’s the first time someone’s mentioned it or brought it up on the show. Very interesting. It’s kind of something that you would go along the lines with just-in-time, it seems like.

Yeah, I think for those not familiar with it, I would use the analogy of your pantry at home. You have your coffee, you have your extra coffee, maybe your extra toilet paper in the garage. As soon as you pull that one out of the pantry, you put it on the shopping list to buy another one. I mean, that’s the same concept. So, you’re probably doing kanban at home, sorry, you just don’t know it with my Trader Joe’s runs every single week. Yeah, exactly.

So, to kind of round out your time at NetApp, and then we’ll transition into some of the things you’re working on now, you also were responsible for implementing, you and your team, a set of standard metrics, specifically around when you were managing high-volume products. So, would you like to have you kind of walk through why you did that, how you did that, and then the impact that it had on the company, because I think standardizing metrics is really important and hard to do.

Yeah, when I took over the global Supply team, we were focused on all of our storage products and anything that was manufactured at a contract manufacturer coming into our assembly or configured order assembly process. The first thing was like we were just talking about, was the kanban, right? So, we said, based on our usage, we want x amount of days. It was three days or four days of inventory based on their replenishment lead time. So, the first thing it did was implement kanban metrics, and we would look at those every week. My team would look at those on every call with their suppliers, and then we’d go over them. If we had any issues, those were things that were highlighted to me, like, “Okay, we’re short on this, we’re gonna… This is a potential risk.” And that led into our ability to meet our on-time delivery metrics. I didn’t necessarily create the on-time delivery metrics, but we definitely lived and breathed by those metrics. And that was the metric that the entire organization was managed to, was 97% on-time delivery. And that one, all of our contract manufacturers were dealing with those metrics. And how did that fare against your company? Was that kind of industry standard?

I think it was. I don’t—I honestly don’t know how that compared. I do know our lead time of roughly three days to turn around a system was definitely a competitive advantage for our company. We had other… I would hear stories from sales teams. They’d go out and then, you know, come back, and we would be in there, and they were scrambling to figure out when they could deliver, and we could have it to you next week, and we’d get the order. So, the lead time was definitely a competitive advantage, and being able to deliver on that, you know, at a 97% rate was something the salespeople could take to the bank.

So, we also… we had other metrics as well, but they’re a lot more detailed, so we can skip those. So, you were at NetApp for a long time, and then you transitioned, and now we’re working at the same company since but in different roles, and it’s kind of a unique situation. So, I would like to have you maybe walk us through and explain what you’re doing now in your current role. Is it currently… I’m back into kind of a commodity manager role. It’s a senior integrated Supply manager. It’s a title I’d never heard of before I got the role, but it’s a category manager or commodity manager role. We are responsible for every aspect of the category. Responsible for the business aspects, the business relationship, supplier relationship, contracts, pricing, supply chain materials, and the strategies around getting the parts from the supplier to the factory or working with the contract manufacturer to make sure they’re building and meeting on-time delivery. So, it’s going back into the playbook a little bit and pulling out a lot of the old plays, but it’s definitely been… it’s been a challenge. We are essentially building the cloud. It’s kind of when people say, “What do you do?” I said, “Well, we’re building a cloud.” Most people are familiar with that term, but essentially we’re building data centers, and we are in a…

I’ve been part of a small team that’s kicked off a new project for AWS that is essentially building our own gear, which is a new adventure for them. And it’s led to a lot of long days and long nights and hard conversations sometimes, but it’s been fun creating and building again. So, you’re kind of like a well-funded startup. We are a very well-funded startup.

So, when you and I were prepping for our interview today, one of the things that really stood out to me is that you said you were a big part of your job is building new supplier relationships. So, why is that such an important part of the job, and what is your strategy for building and maintaining these relationships?

Well, for me, in the well… I learned when I was managing a customer in one of my roles. I learned that there are two ways to manage suppliers. You can beat them over the head and club them over the head every time they do something wrong or every time you want something to get done. And that strategy totally works, you know, it’s a strategy of fear almost, right? If you don’t do it, I’m going to escalate, I’m going to yell at you or something like that, and it is effective. But then there’s the other one of, “I want to be a partner with you, and I’m relying on you to provide me a part, a product, a service, and if you do that for me, I will be successful.” And so I’ve always taken the second approach of, “Let’s be partners. We’re in this together. You’re all I’m only as good as you, and you’re only as good as me. So if I provide you terrible data and terrible communication, how can you provide me good service, good communication, good updates?” I also find that, you know, like I said, beating someone over the head will is effective, but long term, it’s not sustainable. That person is not going to want to go the extra mile for you. They’re going to give you the minimum. When you build that relationship and that partnership, not only do you form an actual relationship with the individual and/or the company and their executive team, but when it comes to, “Hey, I need a favor,” or, “Can you maybe rework those parts for us? You know, I know maybe we gave you the wrong spec and it’s our fault, but could you help us out?” They’re more likely to go do that instead of, “Yeah, sure, I’ll do it, and here’s the bill.” So that’s what I’ve always found is when you build those partnerships, those suppliers will go the extra mile for you. Yeah, and it’s the whole movement from “I” to “we.” We’re a team, we’re collaborating. I think a big part of that is also the innovation piece. Suppliers know their space better than anyone else, and if you can build good rapport and glean insights, you might be able to come up with new ways of building your product or new product ideas that you want coming to you and not to your competitors. Yeah, yeah, and I found too many times, the supplier actually knows your business sometimes better than you because it’s their revenue source, it’s their commission, it’s their lifeblood of their company. So, more and more, anytime they knew my forecast better than I knew it myself, they knew what changed better than I knew it. And when I took over the role of being in that the supplier to an OEM customer, I knew that forecast in and out. I knew how much inventory they had, I knew how fast we were fulfilling it, I knew what was going to be short. And so, when they came and challenged me, I’d go, “No, you’re wrong. That’s not right. That’s not the right forecast you’re running to. It’s right here.”

So, the other thing that you mentioned when we were prepping for today is that you are implementing new processes to ensure supply continuity for your data center build-out. Supply continuity is probably one of the hottest topics in the last three years. So, how are what is your process for this, and how are you building this out to ensure that you’re able to get the supply you need when you need it?

Yeah, I think part of the reason that AWS went down this path of creating this new project was due to supply constraints, capacity constraints with some of their OEM customers, and that became… that was going to slow down their ability to build out data centers and continue to grow. So, they’ve developed this concept. I’m not going to get into too many details, but developed this concept. And, you know, as they brought in experienced supply chain people to lead the effort of building out the supply chain, we’re able to go back and say, “What have we done in the past? How does that apply to what we’re doing today?” You know, trying to set up… we’re not quite there, but trying to set up kanban-type models, setting up flexible models where we are trying to redo packaging. So, we were able to provide more parts in a box instead of just one, enough to build one unit. How about I give you enough to build two units? Re-looking at our supply chain and how we order parts. Instead of ordering in kits, what if I ordered by piece parts? What if I took out the nuts and bolts and screws that I’m putting on a boat and shipping from Europe and sourced those locally? You know, it’s taking a look at all those little things that we’ve been doing and very well, there’s got to be a better way, because at the end of the day, I should never, ever, ever be short a screw, a nut, a bolt, or any, you know, what I like to call “penny” parts or low-dollar parts. They’re just as critical as the high-dollar parts as soon as you run out of them. So, you do not want to hold up, you know, a five-hundred-thousand-dollar system getting out to a customer because you’re short on screws.

Now that’s unacceptable.

So, those are the kind of things we’re looking at, is how do we improve our flexibility as we’re going through this. This is an evolution. We’ve got lots of challenges that we’re working through and some longer supply chains, but we are trying to build and stock finished goods inventory, so as soon as one comes off the shelf, another one can replenish it. And, you know, without an MRP system, it’s hard. We do not have an MRP system. We are running this thing off spreadsheets and engaging with contract manufacturers that do have MRP systems. So, it’s understanding how can we leverage what they have, understanding what that’s going to cost us from material management and so forth. But I think that will be the direction that we’ll probably end up heading over time, is leveraging their systems, min-max triggers, etc., more than we would like to think happens. There are many clients or manufacturers that I talk to or work with that have had a very, very low-cost part or material hold up a very large order. Oh yeah, not all the time. In my previous role, I was managing packaging for Amazon device returns, and, you know, I went from managing five-thousand-dollar storage boxes at NetApp to five-cent pieces of paper and boxes. I mean, and it was… that was a little humbling, you know, thinking, “Well, I used to manage all this spend, now I’m managing this.” And it really wasn’t about that at the end of the day because I realized that, you know, that twenty-five-cent piece of cardboard that goes into the box is going to hold up shipments and our ability to sell that product, doesn’t work, fulfill a warranty requirement to a customer.

Yeah, so it didn’t really matter what the cost was, but you would look at it differently and how you manage it at five cents a piece. Well, I’m just gonna go buy six months’ worth, right? I should never run out of this. And so you set up different processes. Basically, the same. I’m going to set up some sort of vendor-managed inventory or kanban or just-in-time process, but the amount of what I’m buying or the frequency that I’m ordering is going to change based on that. I mean, it comes down as simple as ABC classification-type stuff at that point.

So, you have a career failure that I think is worth sharing because there are a lot of learnings and it just kind of morphed into a success in your career as well. So, I’d like to have you share that with us.

Yeah, it was. That’s a good segue. The last example was managing packaging at Amazon Devices in our Virtual Logistics. I walked into the first day on the job and was basically handed about six spreadsheets for six different fulfillment centers or third-party repair vendors and said, “Here’s our inventory, go replenish it.” You know, no forecasts. I had to go look up BOMs, had to figure out who the suppliers were. It clearly was not going to scale, and we probably had 10 or 20 devices we may have been managing in the beginning, which, as you’ve seen the amount of devices coming out of Amazon Prime Day, it’s multiplied probably 10x since I started there eight years ago. So, it wasn’t scaling, and reverse logistics, as I think we’ve talked before on a previous show, is much more difficult than just manufacturing. Many more variables that you need to take into account. So, I was basically failing. I was running into shortages on nickel parts and 25-cent parts or 75-cent boxes because they were held up in customs in China or at an airport in Chicago, and all of a sudden, we had a pile-up of Echo devices that couldn’t get processed. And, you know, that just shuts down the line. And when that happens during one of your busiest times of the year for reverse logistics, which was January and February after all the Christmas returns, that was a big hit to the ego. Chasing these parts, trying to get them, and it was… the whole process was a failure, and it led to me failing.

So, in terms of how do we turn it into a success, you know, we reached out to our third-party repair vendor who had an MRP system. They had Oracle, and we talked to them about leveraging their tools to build an MRP system and a replenishment system where we could leverage that and then be proactively ordering based on bill of materials, based on forecasts, based on lead time, based on ABC classification.

So, I was ordering small penny parts for six months. Labels, labels, oh my gosh, just give me six months of labels. I never want to talk about labels again, right? Those types of things. And it also, I mean, that was a journey that was not just, “Hey, let’s sit down for a couple of hours and throw this together.” I mean, I was flying from the west coast to the east coast where our supplier was every other week for about three months, meeting with them, whiteboarding, testing, proving this thing out. And we thought we had it nailed, and then we added more BOMs and more products to the database, and it completely blew up because I found out they had done a bunch of hard coding, which is, you know, the ultimate sin in database and Excel, where is our code. So, we basically stripped it down, and then we started all over. And that took another six months of creation, testing, validation, and it was always an iterative process because it was kind of being built on the fly. I’ve learned, and talking to my colleague that I brought over to Amazon with me, he and I worked together on this project after I kicked it off or was in it about six to nine months. He helped me pick that up, and he was much more in tune with databases and things like that, and he kind of helped take it to the next level, right? And what we learned was that there was nothing like what we built in the industry. Oracle didn’t have it, SAP didn’t have it, none of your off-the-shelf cloud MRP systems had this kind of capability that we built. And so, it goes to show how complex it was. And he is now actually rolling out a system to do the same thing that we were doing but trying to scale it across multiple suppliers, and they went to a third party, which is rare in the world of Amazon, as many people know, and he even had to customize his solution. So, I mean, that was, that’s when I was really proud of being able to build it. And the amount of shortages that we had, I didn’t say it went to zero, but it was down to the point of being minuscule and not interrupting production. And we got to the point of actually measuring misses of proactive planning. So, getting out of firefighting, but how many times should you have ordered something and you didn’t, because we got to that level that we could trust the system and take it to the bank, essentially. Not many people get the opportunity to build something from scratch in their career, so pretty interesting.

Well, thanks for coming on our show today, Greg. For those who want to reach out or follow some of the work you’re doing, where would you like to send people to find you?

Definitely reach out to me on LinkedIn under Greg Manhire. If you missed anything, you can check out our show notes. You can find us by typing in “What the Duck?! Another Supply Chain Podcast” in Google. To have optimal search results, make sure to add “Another Supply Chain Podcast” at the end of your search. I’m @SarahScudder on LinkedIn and @Sscudder on Twitter. This brings us to the end of another episode of “What the Duck?! Another Supply Chain Podcast.” I’m your host, Sarah Scudder, and we’ll be back next week.